GenFlightRec/db_import.py

import pandas as pd
import mysql.connector
from mysql.connector import Error
import os
from datetime import datetime, timedelta

# 数据库连接配置
db_config = {
    'host': '152.136.166.253',  # 修改这里，去掉端口号
    'port': 8989,  # 单独指定端口号
    'database': 'fly_ticket',
    'user': 'root',
    'password': 'Cauc@2024'
}

def import_csv_to_db(file_path, cursor):
    df = pd.read_csv(file_path)
    for index, row in df.iterrows():
        sql = """INSERT INTO flight (f_n, f_s_p, f_a_p, f_s_a, f_a_a, f_s_t, f_a_t, f_Date, f_Delay, f_p, f_food, f_wide, f_depcode, f_dstcode)
                 VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)
                 ON DUPLICATE KEY UPDATE 
                 f_s_p = VALUES(f_s_p), 
                 f_a_p = VALUES(f_a_p), 
                 f_s_a = VALUES(f_s_a), 
                 f_a_a = VALUES(f_a_a), 
                 f_s_t = VALUES(f_s_t), 
                 f_a_t = VALUES(f_a_t), 
                 f_Delay = VALUES(f_Delay), 
                 f_p = VALUES(f_p), 
                 f_food = VALUES(f_food), 
                 f_wide = VALUES(f_wide), 
                 f_depcode = VALUES(f_depcode), 
                 f_dstcode = VALUES(f_dstcode);"""
        
        values = (
            row['航班号'],
            row['出发城市'],
            row['到达城市'],
            row['出发机场'],
            row['到达机场'],
            row['出发时间'],
            row['到达时间'],
            row['出发日期'],
            row['出发延误时间'],
            row['economy_origin'],
            row['经济舱餐食信息'],
            row['经济舱座椅间距'],
            row['出发机场三字码'],
            row['到达机场三字码']
        )
        
        cursor.execute(sql, values)

try:
    # 连接到数据库
    conn = mysql.connector.connect(**db_config)
    
    if conn.is_connected():
        cursor = conn.cursor()
        
        # 设置日期范围
        start_date = datetime(2024, 11, 12)
        end_date = datetime(2024, 11, 20)
        current_date = start_date

        while current_date <= end_date:
            folder_name = current_date.strftime("%Y-%m-%d")
            folder_path = os.path.join("D:\college\SE2\Ctrip-Crawler-main\Ctrip-Crawler-withComfortInfo", folder_name, "2024-11-12")
            
            if os.path.exists(folder_path):
                for file_name in os.listdir(folder_path):
                    if file_name.endswith('.csv'):
                        file_path = os.path.join(folder_path, file_name)
                        import_csv_to_db(file_path, cursor)
                        print(f"已导入文件: {file_path}")
            
            current_date += timedelta(days=1)
        
        # 提交更改
        conn.commit()
        print("所有数据成功插入到数据库")

except Error as e:
    print(f"连接数据库时出错: {e}")

finally:
    if 'conn' in locals() and conn.is_connected():
        cursor.close()
        conn.close()
        print("数据库连接已关闭")